Re: [office-formula] Considering IFERROR function: Add IFNA andIFERROR

["David A. Wheeler" <dwheeler@dwheeler.com>]

I said:
> > We could add IFERROR, IFEQ, both, or neither.  I think we should add at _least_ IFERROR. Comments?
> 

Prof. Dr. Andreas J. Guelzow:
> Since IFERROR has been introduced in MS Office 2007, chances are many
> other spreadsheets will implement it to allow spreadsheet import from MS
> Excel. I would therefore agree that we should include that function. 

Actually, how about IFERROR and a new function, IFNA?

Semantics of IFNA(expr, NAvalue) is
IF(ISNA(expr);expr;NAvalue) except expr is evaluated only once.

The problem with IFERROR is that it hides a lot of errors that should NOT be hidden (e.g., division by zero).  NA is a very special error value, so much so that we have special tests like ISNA().  So it'd make sense to have a special IFNA() function as well, just as we have ISNA() that is different from ISERROR() and ISERR().

The IFNA() function would make it easy to create the "FIND" functions the original proposer requested, e.g.:
IFNA(FIND(...);0).

Indeed, the more I think about it, I think IFNA and IFERROR should be in the "Medium" class. They are really useful, and should have been there in the first place.

Note that given IFERROR and IFNA, I think IFEQ is less-needed, and we certainly don't need an IFEQ that has the weird special casing I suggested earlier.

--- David A. Wheeler